Northeastern University Demonstrates Open Source Massive MIMO AI-RAN Prototype

Researchers at Northeastern University's Institute for Intelligent Networked Systems have demonstrated the first open source prototype of a massive MIMO AI-RAN system, announced in a press release. The demonstration integrates ORAN compliant hardware from AmpliTech Group, open source software from OpenAirInterface, and GPU acceleration from NVIDIA into a fully interoperable platform.

The prototype combines an AmpliTech massive MIMO O-RAN Category B radio unit, NVIDIA AI Aerial software for lower RAN layers, and the OpenAirInterface L2+ stack into a single system. The setup uses a two-stage precoding design, with a four-layer MIMO precoder running on NVIDIA’s platform and a 64-antenna beamformer in the radio unit. Tests showed stable throughput across multiple user devices under mobility conditions.

The system validates that a high performance AI-RAN can be built using only open source components. The Northeastern team led system integration and testing, providing a reference design for academic and industry researchers. AmpliTech’s radio unit had previously been certified by Northeastern’s Open6G Open Testing and Integration Center for O-RAN compliance, enabling successful interoperability.

According to project leads, the demonstration confirms that open architectures can deliver both performance and reproducibility, supporting ongoing development of flexible, AI driven wireless networks for future 5G and 6G deployments.